Scan cache guide and usage
- Setting up cache before scan
- Sectors and Tree parameters
- Approximate cache size
- Saving scan cache after scan is complete
- Additional logging and options
- Using previously stored scan cache
- Cleaning the settings
Modern corporate requirements set a very high limits for disk capacity. It's already quite usual to see datastores and iSCSI storages over 20Tb. However disk read speed parameter doesn't improve corresponding to capacity growth. Adding to the picture a common data recovery issues such as bad blocks or RAID corruption cuts disk read speed even more. This makes downtime because of recovery bitter or even unacceptable.
Situation may became even worse if there is a power down or software crashes. Which requires start the scan from scratch.
To deal with such situations Diskinternals have developed and implemented scan cache which saves pointers to file system structures and (if used during repeating scan) allows reading from the scanned disk file system structures only. Using scan cache will not make process instant, as file system structures are to be read anyway. Speed up depends on amount of file system information to be processed but you may expect to get reduction from couple of days to 20 minutes.
Now to the details: there are two methods to set up scan cache. I'd recommend to set up it before scan is started, especially if you expect any issues such as power downs or hardware issues.
Setting up cache before scan
Please click "View->Preferences" to open program options (keyboard short key F11) and press CTRL+E to open Developer tab.
Check "Developer mode" and specify "Write sectors" and "Write Tree" with paths to some existing folder and add a file name. Just make sure it won't write anything to the damaged disk.
Please be advised that scan cache is set independently for each file system. So if you need to scan VMFS disk, you need to set up scan cache in VMFS section.
Normally we place some temporary path at the system disk like: "C:\diskinternals\sectors.cache" and "C:\diskinternals\tree.cache" or "C:\temp\sectors" and "C:\temp\tree". We use C: disk because it's normally a Windows system disk. Actual file names, extension and folder can be changed to any that suits you.
Sectors and Tree parameters
Sectors parameter is a scan cache. It contains only pointers to the disk and doesn't have even a single bit of user information.
“Tree” parameter saves file and folder tree which is required by our developers and engineers to check status of your case. I'd advise to create this log file just in case as it would improve investigation of your case if you would need our support. Creating “Tree” log will not affect performance of scan speed. Data saved in the “Tree” log contains file and folder names plus contents of small files, as contents of very small files (less than 256kb) is stored withing file system contents without real disk space allocation.
Approximate cache size
Scan cache size depends on amount of file system information on the disk. The larger disk you are recovering data from and the more files and data it has, the larger scan cache would be. Normally it is to be expected to be between 1-10Gb.
Saving scan cache after scan is complete
To do this you need to enable Developer mode by clicking "View->Preferences" to open options and press CTRL+E to open Developer tab. Check "Developer mode". Now return to the disk list and right click on the previously scanned disk and select "Save Cache" option.
Please note: should you expect any problems, it would be much better to set up scan before scan rather than hoping it to be finished first.
Additional logging and options
It is possible to set up additional logging about software work at the Developers page. It may be useful if you experience any software crashes. To do this specify a path with file name at the “File Name Log” parameter, check “Additional debug msg” and “Save Log” options.
You may also find useful “Flush” option at the sections corresponding to processed file systems. It will write any new information immediately instead of storing to the buffer. This reduce scan performance but valuable for bad block recognition or if there is some software crash to be fixed.
Using previously stored scan cache
In order to use previously saved scan cache, please reopen Developer tab as described above and specify path to scan cache at the "Read Sectors" parameter in corresponding section. Change “Write Sectors” parameters to use different file name.
If everything was set up correctly, you'll see a following warning message: “Sector cache
Click “Yes” to process scan cache.
Cleaning the settings
Please be advised: scan cache will be used every time disk with corresponding file system is opened. Currently there is no check for different disks and thus once set up scan cache for VMFS(or any other) disk will be applied for any other VMFS disk, which will lead to incorrect results of the scan. To avoid such situation it's important to clear all setting to default, I'd advise to delete “FSM.ini” file at the software's directory.